[10:23] <alan_g> alf_: wouldn't you rather switch USC to C++14 than keep writing "std::chrono::milliseconds"? ;)
[10:28] <alf_> alan_g: yes, it's the plan, but I didn't want to introduce changes to too many aspects of USC. Although I guess that's just a simple switch, so I could enable this now if you would like.
[10:29] <alan_g> I'm not insisting. Just wondering.
[11:08] <alf_> alan_g: thanks for the review, @safe name, do you mean the "com.TestService" or the variable names?
[11:09] <alan_g> the service name
[13:07] <dandrader> What's the expected date for Mir 0.13 release?
[13:12] <alf_> dandrader: No expected date at the moment. We are discussing whether we should release 0.13 or a 0.12.1 bugfix release. Do you need a feature that is going to be in 0.13?
[13:13] <alf_> dandrader: (The discussion about 0.13 vs 0.12.1 is because of the freeze)
[13:13] <dandrader> alf_, yes. this: https://bugs.launchpad.net/mir/+bug/1430315
[13:13] <dandrader> alf_, sounds trivial enough for backporting to a bug-fix release?
[13:15] <alf_> dandrader: possibly, I don't know if it depends on some new internal android backend feature
[13:15] <alf_> kdub: ^^ could we conceivably backport the fix to 0.12 if needed?
[13:16] <kdub> it could just be a new release of the android platform plugin
[13:17] <kdub> but it could also ride along with a 12.1 release easily enough too
[13:31] <dandrader> kdub, should there be a trello card, a bug tag or something else for this backporting to happen, so we don't forget about it?
[13:32] <kdub> camako, what would be the best way to remember for the next release? ^^
[14:17] <alan_g> AlbertA: you're saying NewCallback can't take a lambda? (At least one that doesn't decay to a function pointer.)
[14:21] <AlbertA> alan_g: can it?  but even if it can it has the issue where it won't be deleted if the callback throws
[14:26] <alan_g> Exception safety is not why  you wrote it.
[14:29] <AlbertA> alan_g: sure but it doesn't matter at this point. They are not exception safe
[14:29] <alan_g> ack
[14:30] <AlbertA> In any case, when I looked at their definition they just seemed to take function pointers. I didn't see how I would wedge a lambda but maybe you can.
[14:32] <alan_g> AlbertA: no, you can't unless the lambda decays. Let's write our own test framework...
[14:35] <AlbertA> alan_g: right it can't decay since it captures
[14:39] <AlbertA> alan_g: do you mean RPC framework ? :)
[14:40]  * alan_g was confused